FitRPLS {CPEg} | R Documentation |
performs prediction using Fort and Lambert-Lacroix (2005) RPLS algorithm
FitRPLS(X, y, Xt, yt, LambdaRange = c(0.01, 1), ncompMax = 20)
X |
design matrix, training sample |
y |
binary response vector of 1 and -1 |
Xt |
design matrix, test sample |
yt |
binary response vector of 1 and -1 |
LambdaRange |
tuning parameter |
ncompMax |
number of components to use |
list with two components: misclassification rates and confusion matricies for both the training and test samples.
The columns of the data matrices X and Xt may not be standardized, since standardizing is performed by the function rpls as a preliminary step before the algorithm is run. The procedure described in Fort and Lambert-Lacroix (2005) is used to determine latent components.
MSI and AIM
G. Fort and S. Lambert-Lacroix (2005). Classification using Partial Least Squares with Penalized Logistic Regression, Bioinformatics, vol 21, n 8, 1104-1111.
http://cran.r-project.org/src/contrib/Descriptions/plsgenomics.html
http://www-lmc.imag.fr/lmc-sms/Sophie.Lambert
data(Golub) attach(Golub) #FitRPLS(X, y, Xt, yt)